Analog vs Digital
I’ve been thinking recently about the difference between the analog and the digital, in particular with reference to musical instruments. It seems to me that digital devices often lack a lot of the flexibility of their analog counterparts.
Consider the guitar. To play it you press the strings against certain frets to select a chord and then strum to make a sound. Simple enough? Alright, but you can also slap the strings, or the body, or slide you fingers along the strings, or bend them as you play, or pluck the strings on the tuning head or… a host of other things. In the hands of an expert some quite remarkable sounds can be produced.
Now consider the digital equivalent. There are buttons for each of the chords and a switch to strum. Sure, you can play a bunch of straight tunes, but a lot is missing. You can’t bend it. You can’t slap it. Or rather, you can, but it has no meaning in the digital world. You can do the things the designer intended you to do and that’s about it. And that’s kind of sad.
How do we make digital devices that can be manipulated as organically as an acoustic guitar? That don’t just offer you the choice of 69 different buttons to press, but which offer a rich space of possibilites for exploration. It’s not just a matter of turn buttons into sliders, it has something to do with emergent complexity. To some degree I think it fights against our notions of good design. We don’t want clearly labeled orthogonal functions, we want a complex realm of subtle interactions.
In the case of a guitar we can think of it at two levels. As a beginner we can think in terms of chords that are turned on and off by moving our fingers, and we can produce pleasing results by operating the instrument purely at this level, but it tends to be characterless.
We also have access to the system at a lower level, in terms of the strings and the wood. By straying from the perfect fingering of the abstract level, we can begin to add character to our music. This extra control comes at a cost, there is much more room for error, but the freedom it provides is worth the price. If we want we can ignore the abstract level altogether and play the instrument in completely unconventional ways.
We can’t do the same thing with digital devices because the abstract interaction is the only level available to us. This is considered good design because it makes the device more easier to use and predictable, but it makes interaction feel much more sterile and, well, predictable.
I think the problem is somehow inherent in the nature of a digital device because the interface is kept separate from the implementation. Whereas on a guitar the interface and the implementation are one. Adding more knobs to our controllers isn’t going to change anything if the implementation remains hidden inside a black box.